Photo credit: Aaron Morgan / Quakes Epicenter
2021 has already produced some exciting goals from the San Jose Earthquakes. So much so, in fact, the “team” goals have been lacking this season. If you have to pick between the two, you want more team goals because the data from MLS and leagues we’ve analyzed says that’s how you get more predictable results. I’ve written about this extensively for American Soccer Analysis in my series with Carlon Carpenter called Where Goals Come From.
But, let’s be honest, team goals don’t make Top 10s on SportsCenter each night — BANGERZ do. If you want to create a bit of excitement around a team, sprinkle some of those unlikely goals around the fluid and more consistent team goals. Those unlikely ones from distance are the ones they show on the scoreboard before a game, not the ones that end in a through ball from a 30-pass sequence. The unlikely ones are the ones the social media team will show over and over (and over) again on Twitter and Instagram.
Bangers in the Data
What makes a banger a banger is highly subjective, but we needed to create our own candidate list. We looked to take out the subjectivity, although probably not the debate, and we found a way to (mostly) analyze these kinds of goals with data.
What makes a banger? We spent some time playing with some of the metrics that are often used to identify top players and goalkeepers, and used them for a different purpose. Using the data from our American Soccer Analysis project, we pulled over 58,000 shots from 2015 to 2021 — effectively what is known as the TAM Era.
We also decided not to use the ASA xG model, as it is geared more towards team and season predictiveness. Instead, we needed a model which is geared toward the highest shot evaluation accuracy. The ASA model uses about about 20 different data points of a shot. In our Where Goals Come From xG model, we have expanded this to 42 different data points for each shot. We included shot data from the European Big Five leagues to give us a model with 350,000 shots. We provided a league flag to the model so it can find differences between the leagues, although it’s not as important as you might think. Not content to rest on our laurels, we then created an improved hybrid model for MLS by making a second model from just the 58,000 MLS shots, adding the two values together and dividing by two. Although every model has its blind spots, which the Where Goals Come From project will describe in a future ASA article, this approach really gives the best perspective on each particular shot that is possible with the data currently available to ASA.
In addition, we took the shots on target data (saves + goals only) and created what’s called a “post-shot xG” (PSxG) model also called a “keeper model”. The typical use of a PSxG model is to evaluate the difficulty and chance of a save for a perfectly-positioned goalkeeper (using some fancy geometry that ASA has confirmed with some goalkeepers and coaches). A perfectly-placed shot will usually have a high PSxG value. A shot that gives the model a set of data points that are highly unlikely to produce a goal will have a low xG. Thereby, the best-of-the-best banger goals will have a low xG and a high PSxG.
We then looked across the 6,400 MLS goals from the 58,000 shots to determine the right criteria for a banger goal. After some experimentation, we set the cutoff at 0.05 xG or a shot with a 5% or less chance of resulting in a goal based on our 42 factors. We also felt that the shot didn’t need to be outside the 18-yard box, given the box does not have equal dimensions, but that 18 yards was a fair minimum distance. Now, we didn’t eliminate free kicks, but to be honest, they average around a 6% chance of scoring which makes them unlikely, but not impossible, candidates for our list. It is well-established that open play goals from distance are more difficult to score than their dead ball counterparts, and that shows up in the xG modeling.
Lastly, we want to make sure we only got shots on the edges of the goal. Given this thing called gravity, a surprising number of shots do not end up in the top 10% of the goal. “Top bins” or “upper V” is not a myth, but goals in these areas just aren’t as common as we think they are.
Using these criteria, out of the 6,400 goals in MLS we identified 285 bangers (that’s about one every eight games), and the Earthquakes are tied for ninth with 13 of them. Now we can get to a top-five list, except there is one problem: these 13 goals are very much spread across the spectrum of our xG and PSxG with no clear-cut winner (I’ll share the chart at the very end — no looking ahead!). That means we’re going to need to use a bit of judgment in order to arrive at a top-five list and declare a winner.
Okay, enough of the technical stuff. Let’s get to the goals themselves.
Honorable Mentions
To start with, there are some goals that are not “top-five” but definitely merit mentioning. We’ll call these the Honorable Mentions.
Anibal Godoy – March 11, 2017 vs. Vancouver Whitecaps
Distance: 33 yds, WGCF xG: 0.025, ASA xG: 0.021, Post-shot xG: 0.16
First off, this is almost assuredly not the goal you thought it was. The Anibal Godoy chip you are probably thinking of was the week before this goal on March 5th against the Montreal Impact. This goal came against a 10-man Vancouver side as the Quakes were probing for a way to play through.
As was typical for Jahmir Hyka, his dribble from a static position went nowhere, and he simply passed back to Godoy who hit a 33-yard curler to the back corner. This goal was the game-winner, and the Quakes secured six points from their first two games in 2017 from two Anibal Godoy goals. Makes you wonder why we didn’t get more of them from Godoy, huh? (Psst, it’s because bangers are rare.)
Cade Cowell – August 29, 2020 vs. LA Galaxy
Distance: 29 yds, WGCF xG: 0.026, ASA xG: 0.037, Post-shot xG: 0.21
Many will say that David Bingham should have done better here. Indeed, it’s for that reason the post-shot xG says 0.21. That means a keeper in the right position should be able to save it 79% of the time. Don’t blame Cade, though, he did all the right things. This was quite a strike for the then 16-year-old in his very first MLS start, and the Quakes temporarily took the lead on the road in this one, before coughing up two.
All Quakes fans were hoping this is the first of dozens of more goals to come. So far, so good.
Eduardo “La Chofis” Lopez – May 22, 2021 vs. Sporting KC
Distance: 23.5 yds, WGCF xG: 0.037, ASA xG: 0.032, Post-shot xG: 0.60
This won’t be the last time Tim Melia gets picked on with near-perfect placement in this article. This back-post curler from Chofis is almost not even saveable regardless of where Melia is positioned. The touch is so light, it just shows that you don’t need a rocket: you need placement.
The corner kick here is a designed play taken “off the training ground”, as Matias Almeyda mentioned in the postgame press conference, and it’s difficult to imagine Chofis or anyone doing better with it. The Earthquakes grabbed an early 1-0 lead here, but ended up losing 1-3 in this one.
Our Top Five Bangers
Let’s count down the top five bangers of the TAM Era for the San Jose Earthquakes.
Coming in at #5: The Chip
#5: Quincy Amarikwa – March 13, 2016 vs. Portland Timbers
Distance: 35.5 yds, WGCF xG: 0.042, ASA xG: 0.193, Post-shot xG: 0.48
Let the debates begin, because there is no doubt this goal from Quincy Amarikwa should be on this list, the question many will ask is “why isn’t it #1?” Well, that’s a fair question. It is the longest goal on the list. The placement is about as good as possible, resulting in a 0.48 post-shot xG, but that’s a bit of the problem — the keeper’s positioning is…not great. I mean I had to look up who the keeper (Adam Larsen Kwarasey) even was, so that tells you something.
Take nothing from Amarikwa, not only did he win the ball and take on three players in a dead sprint, it’s the goal of his career, and one of the most memorable in the history of the Earthquakes. The xG models can’t agree on this one, but the WGCF model (with twice the data points) has it twice as high as ASA (and higher than all the others in our list), and the WGCF model is the one we are riding with here. Once again, The Chip is robbed of the top billing Quincy and his Chip Apologists will feel it deserves. Sorry, Quincy.
Coming in strong at #4, it’s the Magnus Eriksson goal you might have forgotten about.
#4: Magnus Eriksson – August 29, 2018 vs. FC Dallas
Distance: 27.5 yds, WGCF xG: 0.037, ASA xG: 0.04, Post-shot xG: 0.71
This shot is virtually unsaveable. With a super-high post-shot xG of 0.71, this Magnus Eriksson wondergoal is one of the best individual efforts from the worst season in team history against the team that couldn’t figure them out, FC Dallas. With the game tied at two, Magnus passes to Wondo inside the box against a packed FC Dallas defense and finds himself with just enough space when receiving the ball back to take a touch to his left and sweep a curler to the back post (notice a pattern here?). It’s the best shot of Eriksson’s Quakes career and at only 0.035 xG, it squeaks past Quincy to earn our #4 spot.
At #3 is the top set-piece goal from the only player with two goals on our qualifying list.
#3: Nick Lima – July 29, 2017 vs. Colorado Rapids
Distance: 30 yds, WGCF xG: 0.031, ASA xG: 0.029, Post-shot xG: 0.54
This goal wins the prize for being both low xG and high post-shot xG. We didn’t know that Nick Lima could hit a ball like this until he did it to beat the Rapids 1-0 on the day. Whether purposeful or not, the corner kick glances off the head of Darwin Ceren, but ends up at the feet of Lima who calmly knocks it down, lets it bounce once, and just rifles it into the top of the near post. It’s tough to imagine this goal being saved, and, for a shot to the near post, it probably gets the best possible post-shot xG it can muster at 0.54.
Lima would score a second banger that met our criteria the following season with a left-footed, far-post curler on the road against the Vancouver Whitecaps.
And now it gets interesting, because it’s a battle between our lowest xG goal and our highest post-shot xG goal and between two recently-rejuvenated players on the roster.
Taking the runner-up spot…
#2: Paul Marie – April 16, 2021 vs. Houston Dynamo
Distance: 34 yds, WGCF xG: 0.017, ASA xG: 0.017, Post-shot xG: 0.21
Paul Marie? Paul Marie! This goal is similar to Cade’s and another banger that didn’t quite make our list from Jackson Yueill in Houston 2 years and 3 days earlier, except it is from way…down…town…BANG! (An old ESPN SportsCenter reference there.) If there was a 2-point shot in soccer, it would have been behind the line.
This “curling, whirling durbish” (as the Houston announcer said) was about seven feet shy of Quincy’s — that’s how far out it was. The criticism with all these kinds of goals is the late break by the goalkeeper and/or the questionable positioning. The same criticisms against Bingham can be leveled here on Marko Maric. The post-shot xG says it had a 79% chance of being saved. But again, it is hit with such pace and such bend (which the ASA data does not know about) that it is probably a good bit less than that.
What is with the Quakes and amazing game-one-of-the-season goals? First there was Godoy’s chip in 2017, then the 20+ pass goal finished off by Hoesen in 2018, followed by Alanis’ free kick to draw with TFC in 2020, and now Marie’s absolute banger in 2021.
Marie’s was a near-perfect goal, and it is the lowest xG of all of the Quakes’ goals since 2015. But if you are going to be numero uno on our countdown — if you are going to be proclaimed King of the Bangerz — your goal has to be p-e-r-f-e-c-t. Our #1 banger is just that.
#1: Shea Salinas – April 20, 2019 vs. Sporting Kansas City
Distance: 24 yds, WGCF xG: 0.039, ASA xG: 0.034, Post-shot xG: 0.84
Ah, the Salinassance was an amazing run in early 2019. Shea “Messi” Salinas had a stretch of four goals in four games, and this was his best. For pure scoring technique, the Cali Clasico ones can’t top it. Some of those were more important (although the Quakes at 1-4-0 at this point in 2019 needed a win badly), but this goal is absolute perfection. The cut inside, the angle, the placement, everything about it is *chef’s kiss*.
It is the very definition of an unsaveable goal. It has the best post-shot xG for a goal under 0.05 xG in the entire ASA dataset — it other words, it was the least saveable banger in the league since 2015.
Shea looks like he might just be trying to chip a pass to the back post. You know that whole coachspeak of “pass the ball into the goal”? Salinas took that to heart here. Inch-perfect crossing technique can also be used for inch-perfect goal-scoring technique. Who knew?
In addition, it happened 10 seconds after kickoff. Want to just bury your opposition’s will? Do THAT.
Second fiddle to Wondo no more, our King of the Bangerz — Robert O’Shea Salinas!
Hey, Shea, Jean-Baptiste Pierrazi just called from 2014, and he’d like a word.
